Also, have you tried micro-plasma arc welding instead for something as thin as what you're welding?
Another often overlooked requirement that's necessary for that type of work is to make absolutely sure everything in the welding location is squeaky clean room clean!!!
I'm talking about the material is checked to be clean of any residue, the tools, torch, tungsten grinder, wire brushes, etc. are dedicated for those specific metal alloys...
I would imagine that the welds are autogenous correct? If so, the wiping the diaphragms prior to welding with a lint free cloth only to be used with that particular run of joints and replaced when starting a new run every day... Have you also checked your torch parts to make absolutely sure that they are also squeaky clean? This must be done religiously on an hourly basis an I cannot emphasize this enough!!! Even the welding gloves, hand tools and anything else one can think of must be clean and without any potential residue contaminating the metal... If you are using a filler wire, then make sure that there's no oxidation or residue on the surface of the wire... But I doubt that you are.
The Hastelloy can be a bit tricky to weld for some but, once they understand the characteristics and gain sufficient practice the welder usually produces sound welds...
I agree with 46.00 in objecting the need for H2 gas mix for such thin material but more importantly IMHO. is why you're not welding these with the micro-plasma process variant instead? For one thing, the arc column would be more constricted so that the electrons are being focused more on the joint location, and could amount to better penetration with less heat input and limiting the size of the HAZ... Of course there's also no need to worry about possible tungsten contamination by making contact with the joint because the electrode doesn't stick out from the torch and isn't exposed and one can also use a greater torch to work distance than with GTAW...
Also check your gas purity as well as your purging gas purity if used too because if your gas isn't up to snuff or one of your hoses are leaking could result in having contaminated gas or a mix or the gas and the air being sucked in from the leak... With that type of work, I would demand at least medical grade gas instead and the cylinder to be dedicated unless you have manifold system which would mean that you would need to check out that system as well... All of these factors must be examined and each potential aspect for contamination must be eliminated in a methodical manner as well as anything else that may contribute to deficient welds and lower your failure rate substantially because frankly speaking, if everything was in optimal condition, and there's no pointing fingers at the welders skills, you should have less than a 5% reject rate and that should be the goal you want to reach initially... This is very much so achievable!!!
Respectfully,
Henry